A Reflexive Inquiry on the Effect of Place on Research Interviews Conducted With Homeless and Vulnerably Housed Individuals

Abstract: In this study, I utilized a process of reflexivity to examine the effect of location when conducting interviews with homeless and vulnerably housed individuals. The impact of interview locations has received limited attention in the community psychology literature, despite the majority of research being community-based. The study provides insights into the challenges, benefits, and power relations involved in selecting a research interview site and in conducting interviews. Personal journal entries were used to analyze the effect of location on the participants and I as the researcher, through a comparative analysis of interviews conducted in the community and a research center. Results demonstrate that interview locations hold great amounts of power and can provide the opportunity for holistic understandings of research topics. Lessons learned and methodological implications are discussed.

Key words: interview sites; interview location; homeless and vulnerably housed participants; reflexivity

2.2 The impact of location on interviews with homeless and vulnerably housed individuals

Community-based research is a dynamic process that can place a researcher into a wide range of settings. Personally, flexibility has been a major tenet of my methodological philosophy. When planning the site for interviews and focus groups, a main consideration is the needs of the participant (e.g., accessibility issues) and not necessarily my own. As a result, I have been taken outside the halls of academia and have conducted interviews in hospitals, parks, community agencies, homeless shelters and people's homes. [1]

The importance of interview location gained personal significance when my small research center began the process of moving to a larger building. Would participants be as receptive to our new location? Would they feel as comfortable entering a building with complicated hallways and stairways? These questions led me to reflect upon what constitutes a good interview location. Do interviews conducted in the field differ from interviews conducted at research centers in academic institutions? What are the power dynamics involved in deciding an interview location between "researcher" and "participant?" [2]

This article addresses the importance of interview locations for research interviewers, particularly when conducting research with individuals experiencing homelessness and vulnerable states of housing. It focuses on how interview locations can impact the research process, the interviewer, and the interviewee. I begin with a review of the impact of interview location. Following the literature review, my experiences as an interviewer for a mostly quantitative project with homeless and vulnerably housed individuals is described via a personal research journal. A brief comparison of data quality among participants who completed the surveys in the community to those who completed the surveys at our research center is also presented. In conclusion, lessons learned from the field are discussed with a particular emphasis on the implications of selecting a research interview site on the research process, the interviewer, and the interviewee. [3]

Conducting research with individuals experiencing homelessness or vulnerable states of housing can pose some challenges. Homeless and vulnerably housed individuals can be difficult to locate (MOWBRAY, COHEN & BYBEE, 1993) and gaining access to the population often requires a negotiation with social service agencies (TAYLOR, 1993). Once access to the population is acquired, potential research participants may not be prepared to engage in an interview due to psychological impairments or substance use (VANCE, 1995) or distrust of the interviewer (ROSENTHAL, 1991). Others who have exited homelessness may wish to avoid reflecting on past problems and choose not to participate (MOWBRAY et al., 1993). [4]

When individuals do agree to participate, several considerations require attention. Potential participants may not show up at the scheduled interview time (HWANG et al., 2011). For housed individuals, physical housing barriers may exist, such as missing buzzers or inaccurate buzzers (ibid.). For individuals residing in shelters, interview times are contingent upon operating hours (ibid; see also TAYLOR, 1993). Despite these challenges, there are other important elements to the research process. One of the main goals of research with marginalized groups is to provide a "voice" to participants (CLOKE, COOKE, CURSONS, MILBOURNE & WIDDOWFIELD, 2000). In speaking with homeless young people, ENSIGN (2006) found that the majority of her participants reported positive experiences when involved in research projects. They particularly enjoyed being able to tell their stories to the researchers and have them listen. [5]

2.2 The impact of location on interviews with homeless and vulnerably housed individuals

For research conducted with individuals experiencing homelessness and vulnerable housing, the selection of an interview site requires considerable thought. There is a rich literature in the field of ethnography that has discussed the diverse settings that researchers place themselves when conducting research with individuals experiencing homelessness and vulnerable housing (e.g., CLOKE, MAY & JOHNSEN, 2010). CLOKE et al. (2010) argue that the aim of ethnographic work with individuals experiencing homelessness is to present the participants in their true identity, capturing their attitudes, experiences and, inevitably, an understanding of their living situation. In other disciplines, the same attention has not been paid to details regarding the location of interviews with individuals experiencing homelessness and vulnerable housing. Interview locations should be determined based upon the preferences of participants; however, this may not always be feasible (COHEN et al., 1993). For example, ROSENTHAL (1991) describes interviewing individuals in parks, coffee shops, cars, boxcars, and so forth. TORO (2006) recounts sampling individuals from parks, bus stations, and other public places. These types of scenarios bring about ethical and safety concerns. Ethically, public spaces do not guarantee privacy for the interviewee. As the information that is shared during interviews is often personal, a public environment can potentially nullify the right to confidentiality for participants. From a safety perspective, open spaces can make it difficult for researchers to scan areas for potential vulnerabilities. [6]

Interviews conducted in an individual's residence bring about challenges different than those conducted in public spaces. Interviews in private residences can mitigate the ethical issues described above, since outsiders are removed from the situation. Safety issues remain, however, as interviewers can be placed in potentially precarious situations. Ways to promote the safety of interviewers include sending multiple interviewers to a private interview site, ensuring interviewers have cell phones, and providing interview training (ibid.). These types of steps are important, as some participants who are vulnerably housed will prefer to have the interview completed at their home (COHEN et al., 1993; TORO, 2006). It is important to be open to home interviews, as the interviewers in the COHEN et al. (1993) study felt that interviews where the participants had control (e.g., their own home or apartment) differed from those where participants did not have control (e.g., a group home, a relative's home). [7]

Quantitative evidence for interview differences based upon location has been completed in non-homeless samples. For example, among a sample of adolescents being surveyed on health risk behaviors, KANN, BRENER, WARREN, COLLINS and GIOVINO (2002) found that surveys completed at school produced estimates indicating higher risk behaviors (particularly illegal and socially stigmatized activities) than those surveys completed at home. Interviews conducted in the home also pose an interesting dynamic for the researcher. JORDAN (2006) details her experience as an ethnographic researcher in the homes of families. She notes that the home is an intimate setting for a researcher to enter and one that includes physical and psychological boundaries. Due to these boundaries, researchers must negotiate their simultaneous role as guest and investigator. The researcher also gives up some form of control in the research process, since they are not in ownership of the setting. [8]

The idea of "control" mentioned above has been noticeably absent from the discussions of interview locations with participants who are homeless and vulnerably housed. The power dynamics among researchers and research participants are important considerations, particularly for community-based research with marginalized groups. The relationship between research investigator and participant can impact the quality of the data that is collected (KELLY, 1990). Therefore, the research process should involve a collaborative framework between investigator and participant (ibid.); however, due to the inflexibility of the scientific method, collaboration is often difficult to achieve. The positivist-informed scientific method enables control of the research process by the researcher (BRODSKY, 2001; WALSH, 1987) and strives to limit subjectivity inquiry (NEWBROUGH, 1992). As a result, and particularly in quantitative-based methods, research participants are viewed as subjects under the control of the research demands and the researcher, and the imbalance of power among the researcher and the participant is promoted. This power imbalance can impact the selection of an interview location, as laboratory-based research is viewed as superior to community-based methods (KELLY, 1990; NEWBROUGH, 1992) therefore limiting the opportunity for alternative interview locations to be considered. [9]

The rigidity of most laboratory-based research is challenging for research with individuals experiencing homelessness and vulnerable housing. The daily experience of being homeless or vulnerably housed may make it difficult for some individuals to maintain scheduled research appointments (HWANG et al., 2011) or acquire the necessary resources (i.e., transportation) to get to interview sites on university campuses. Individuals experiencing homelessness can also be very difficult to locate for research studies (MOWBRAY et al., 1993) which makes flexibility in the research process imperative. For example, even difficult to locate participants may access services, such as meal programs or drop-ins, so researchers may need to go to these services to locate individuals and potentially conduct interviews on the spot. Relying solely on campus-based laboratories will therefore limit the opportunity to interview individuals experiencing homelessness and vulnerable housing. Additionally, since individuals who experience homelessness and vulnerable housing are marginalized populations, the power differentials inherent in a laboratory-based setting may further enhance feelings of marginalization. Providing a choice in interview location may be a small opportunity to empower participants in the research process. Although this falls short of true participatory action research, where participants of the research are actively involved in its inception, delivery and interpretation, asking participants where they would like the interview to take place establishes some form of participation in the research methodology. [10]

The choice of interview location not only impacts interviewers and interviewees, but can also provide meaningful data before the interview even takes place. When choice is involved in determining the interview location, ELWOOD and MARTIN (2000) encourage researchers to reflect upon the location choice in terms of its meaning to the community and the individual. Researchers should also consider how this choice can influence data collection and the interview experience. Available interview locations, proposed by both the interviewer and interviewee, and their selection can be indicative of the importance of certain locations, highlight a lack of available locations, and provide an understanding of potential social divisions within the community (ibid.). The choice of location and who decides on the location is also an important consideration in the power dynamics involved between the interviewer and the interviewee. Certain locations may situate participants differently in terms of their responses and contribution to the research (e.g., interview conducted in an area familiar with the participant opposed to one that they are unfamiliar). [11]

This review has demonstrated that the selection of locations for interviews conducted with individuals experiencing homelessness and vulnerable housing is a layered process that has largely been ignored within the literature. Much of the literature cited is close to twenty years old, indicating that an updated examination is required. [12]

This article describes my experiences as an interviewer for a mostly quantitative study focusing on homeless and vulnerably housed individuals. It is based upon my reflections into the importance of interview location for research conducted with individuals experiencing homelessness and vulnerable housing. Through reflexive journaling, I detail the impact of interview locations on the research process and me as the researcher, and my interpretation of the impact on the research participants. I also include a comparison of participant responses and subjective indicators (e.g., rating of interview quality and rating of participant interest in the study) for interviews conducted in the research center to those conducted in the field. I conclude with lessons learned from this process and future directions for methodological considerations when beginning research with individuals who experience homelessness and vulnerable housing. [13]

This article provides a novel contribution to the field and in particular, FQS. The topic of reflexivity has been a recurring theme throughout the journal, as well as the relationship between qualitative and quantitative research. What has yet to be investigated within the journal is the application of a reflexive methodology when conducting research with individuals experiencing homelessness and vulnerable housing. As discussed above, the power dynamics between researcher and participant is potentially exacerbated when working with individuals experiencing homelessness and vulnerable housing. Therefore, engaging in reflexivity, regardless of whether the study is quantitative, qualitative, or mixed methods, is particularly important when conducting research with this population. There have been few examinations of the impact of interview locations, regardless of the study population being investigated. Discussions of location arise in reflexive work, however few explicit examples exist in the literature. [14]

This section outlines the research project in which I served as an interviewer, a brief description of who I am, the data collection process, and data analysis strategy. [15]

The Health and Housing in Transition (HHiT) study is a multi-site study conducted in three urban cities in Canada. The longitudinal project followed approximately 1,200 homeless and vulnerably housed individuals. The current research focuses exclusively on one of the sites. The sample included men and women over the age of 18 who were recruited from homeless shelters, meal programs, and rooming houses or similar type of marginal housing. The HHiT study received ethical approval by the research ethics boards at each of the three sites. Participants were compensated $20 for their participation and bus tickets were provided to participants when the interview was conducted at the research center. [16]

The survey protocol was largely quantitative, with a few open-ended questions scattered throughout. Participants were asked a series of questions regarding their housing, health, substance use, quality of life, service use, social support, and community integration. The open-ended questions asked participants to describe what they liked best and least about their housing and neighborhoods. Participants were also asked how they think their housing impacts their health. At the end of the interview, interviewers were asked to rate the interest and co-operation of the participant on a 1 to 5 rating scale (1= not at all to 5=extremely interested and co-operative) and how much difficulty the participant had answering the questions on a 1 to 5 rating scale (1=not at all; 5=extreme difficulty). The interviews took approximately 45 to 60 minutes to complete. [17]

This article is a byproduct from a largely quantitative study. The article resulted from reflexive thinking on the limitations of a quantitative design when considering methodological impacts on the research process. This article demonstrates the utility and relative ease of including reflexivity within quantitative designs. It is in line with previous research published in FQS, but demonstrates the particular importance of its application when conducting research with individuals experiencing homelessness and vulnerable housing. [18]

I am a white, English-speaking, male who was in his late-20s during the time of this study. I was a graduate student in experimental psychology, social-community stream, at a university located in one of the study sites. The social-community stream of the experimental psychology program focused on applied social research that was values-based and dedicated to working with vulnerable populations. My dissertation research used data from the HHiT study and I became involved with data collection at the third stage, or the second follow-up year. I had previous experience conducting research with marginalized populations. [19]

During the third year of data collection, and my first year of involvement in the study, I began to reflect upon whether my interviews differed in quality depending upon the interview site. I had been to several different locations to conduct interviews, including the research center (which served as the administrative hub for the study), homeless shelters, drop-in centers, and individuals' homes. As each of these sites were quite different, I was curious as to how each interview location brought something new to the interview process. As there were several interviewers hired for the study, I shared these thoughts with another one of the interviewers and we began a dialogue on the topic that also helped to frame my thinking. [20]

Based upon my reflections into the importance of interview location, I began a process of reflexivity. Reflexivity is a common practice within qualitative inquiries, however it has received little attention within quantitative studies (RYAN & GOLDEN, 2006). RYAN and GOLDEN further state that discussions of who, where, and how the data is collected is often superficially described in quantitative papers. Therefore, by including a reflexive component in this largely quantitative survey, it will provide new insights into the process of reflexivity in quantitative research conducted with homeless and vulnerably housed individuals. [21]

Definitions vary, but reflexivity can be loosely defined as an, "explicit, self-aware meta-analysis of the research process" (FINLAY, 2002, p.531). It recognizes the subjectivity involved in research process and challenges researchers to locate themselves in data collection and analysis (HALL & CALLERY, 2001). Researchers face the intersections of the research process, the participants, their own lives, and the outcomes of research (ST. LOUIS & CALABRESE BARTON, 2002). Therefore, reflexivity is a process that continues throughout the research. It is an important process, since researchers hold inherent biases within their characteristics, roles, thoughts, perceptions, and practices (BREUER & ROTH, 2003). [22]

One method of reflexivity involves autoethnography (DOWLING, 2006). This involves the use of autobiography and narrative inquiry via personal research diaries. Through autoethnography, the researcher examines one's own personal values through a social context or how we are engaging among the lives of our research participants (ALSOP, 2002). Research diaries can act as a response to positivistic research methodologies, since they acknowledge that researchers will have different assumptions about the world (ELLIS, ADAMS & BOCHNER, 2011). Research diaries can also help a researcher to explore methodological issues within their studies, add to the interview data, aid in the organization of the study, and allow the researcher to reflect on his or her interviewing style (NADIN & CASSELL, 2006). As research diaries have been used for similar analyses (e.g., CLOKE et al., 2000), it was determined that this method was the most fitting for the current reflexive analysis. [23]

Before the start of the third year of data collection, I began journaling my experiences after each individual interview in order to capture the subjective influence of interview location. The research diary not only served as a means for personal reflection, but it also allowed for an ethnographic analysis to be conducted, something that extended past the goals of the HHiT project. Without having strict limits as to what to journal, I decided to reflect upon: 1. assessment of the location; 2. the impact of location on the quality of the interview; 3. the comfort of the interviewer and interviewee; and 4. general comments. Journaling would ideally take place immediately after the interview had finished, but sometimes due to time constraints this immediate journaling did not occur; however, journaling was always completed the same day of the interview. [24]

Journal entries were conducted for a total of 43 interviews. Some interviews were excluded due to researcher inability to journal (e.g., time constraints) or interview type (e.g., phone interview). The length of the interviews varied from 30 to 90 minutes. The gender of participants was skewed based upon my gender, as I predominantly interviewed male participants with the exception of three females. It was thought that female participants may not feel comfortable answering some of the more sensitive questions asked during the interview with me. Interview locations were varied and included the research center, community drop-ins, parks, homeless shelters, a hospital, and an individual's home. [25]

The journal entries were compiled into one document. All journal entries were labeled with the date, the interview location, and the gender of the participant. I reviewed this document before any coding was conducted. This was undertaken to become reacquainted with interviews that had been conducted months prior. After this initial review, I coded the journal entries. The data was analyzed using a general inductive approach, through a series of steps, specifically, first cycle and second cycle (SALDANA, 2009). In first cycle coding, open coding of the data occurred and preliminary codes were developed. Each response was read line-by-line and codes were developed for segments of the data. As initial codes should stick closely to the data (CHARMAZ, 2006), in vivo coding was used as often as possible. Following open coding, second cycle coding was completed. This type of coding allows data to be synthesized and placed into meaningful categories and subcategories. During this stage of coding, a constant comparison technique was used. Within this, codes and categories were compared within each individual journal entry and then across all of the journal entries. Regular reviews of the data can correct fatigue by the analyst and maintain consistency throughout the process. [26]

Following the coding process, I shared my coded document with another interviewer who had discussed the topic with me. This discussion served as a peer review function, as well as a reflective opportunity. The discussion led to the verification of the themes that were developed. During the analysis process, the journal entries were continuously referenced and reviewed to ensure accuracy of the codes and themes. The final list of themes was reviewed several times to ensure accuracy. [27]

It should be noted that throughout the research process, I was in regular contact with the other interviewer and discussions often emerged about how our interviews were going. This served as a time of reflection and debriefing for both of us, but also undoubtedly improved or enhanced my journaling process. Through regular contact, we were able to learn from the other's experiences and think about issues that we may not have previously considered. As systematically considering the impact of location was a new process, these interactions strengthened the data collection and data analysis methods. [28]

Descriptive statistics were compiled for the participants on the interviewer ratings of the participants' interest and difficulty with the interview. Participants who had their interviews conducted in the field were compared to those who had their interviews conducted in the community. [29]

Twenty-eight of the interviews were conducted at the research center and fifteen were conducted in the community. Community interview locations included homeless shelters, drop-in programs, parks, a coffee shop and the participants' own housing. The housing situations of the participants slightly differed. For participants who conducted their interviews at the research center, close to two-thirds had their own apartment or lived in a rooming house while the remainder were either homeless or temporarily staying with friends or relatives. For participants who conducted their interviews in the community, 80% were living in their own housing or in a rooming house and the remainder were either homeless or temporarily staying with friends or relatives. As many of the community interviews were conducted in the participants' homes, these findings make sense. [30]

As noted below in the reflexive findings, not all participants were offered a choice in selecting an interview location. A total of four participants were not asked where they would like the interview to take place. Twelve interviews were unplanned, nine occurring at the research center and three occurring in the community. These unplanned interviews were the result of participants showing up to the research center without an appointment or when I visited the participants' places of residences to try and establish contact. [31]

The interviews conducted in the community were rated more positively than those conducted at the research center. All of the participants who completed the interview in the community were rated as being very or extremely interested and cooperative during the interview. For interviews conducted in the research center, over 75% of participants were rated the same. In terms of difficulty with the interview, all of the community participants and 93% of research center participants had no difficulty or less than average difficulty. The percentage of missing data for interviews conducted in the research center did not differ from those interviews conducted in the community. [32]

Results are separated into three sections and report on my personal reflections on the influence of location on: 1. the interview process; 2. the interviewee; and 3. the interviewer. Within each section, the interview location site (e.g., in the research center or the community) is compared and contrasted. The headings presented below each section heading represent the main themes to emerge from the data analysis. [33]

For interviews conducted at the research center, I felt a great power imbalance between myself and the participants. This imbalance greatly favored me, due to economic and class-based privilege and the engagement with the participants was predominately one-way. For example, early journal entries included such passages as, "I don't think place affected our interview that much ... He gave honest responses and elaborated when necessary," and "He was forthcoming with his responses and the location did not have an effect on the interview." These quotes highlight that I did not think interview locations held much power, but this was under the pretense that the interviewees were providing me with their responses. My quotes indicated I had a belief that interview location was redundant as long as the interview went according to plan. In this way, conducting the interviews in the research center perpetuated a traditional, positivistic research approach. Therefore, the research process followed a very traditional trajectory in that I was in control, dictated the tone of the interview, and acquired the necessary survey answers from the participants. [34]

For interviews conducted in the field, I felt greater equality between myself and the participants. Although the power imbalance was always in my favor, due to the previously mentioned economic, class-based privilege I had, I felt interviews conducted in participants' own homes or locations in the community (e.g., a shelter, community center, a park) lessened the imbalance. This was particularly evident during home interviews, where I held a position other than just a researcher - I was now a guest in someone's home. Not only was I concerned with interviewing participants, but I also had to be aware of the social conventions of being in another's home. This may have served as a distraction and could have impacted my delivery of survey questions or opportunities to probe for more detailed responses so as not to come off as an intrusive guest. [35]

Therefore, in my interpretation, community interviews involved two power dynamics—one as researcher, where power was maintained, but also one as guest, where power was slightly relinquished. These competing influences lessened my control of the interviews, particularly in comparison to the interviews conducted in the research center. These setting-dependent power dynamics may have influenced the research process. At the research center, I felt in control and guided participants to answer the survey questions. In the community, I was less in control and had to contend with multiple identities (i.e., researcher and guest) making it potentially more challenging to elicit the same standard of interview as in the research center. [36]

Interviews conducted at the research center involved greater situational control than interviews conducted in the field. In the research center, interviews were conducted in a private room and distractions were minimal. In contrast, some challenges were encountered during interviews conducted in open-air spaces. However when interviews were conducted in the home, there were instances where other residents interrupted the interview. In other cases, there was the potential for other individuals to enter. For example, in one journal entry I state, "We conducted the interview in the participant's kitchen, which he shared with the other tenants in the building. No one came in while we were doing the interview." As a result of these examples, I placed greater emphasis on ensuring the participants' confidentiality and halted interviews when other parties were in close proximity. Interviews conducted in community settings (e.g., homeless shelters, day programs, parks) were also subject to similar issues as home interviews. Therefore, community interviews involved a greater potential for unwanted distractions which led to more disjointed and lengthy interviews, invariably affecting the interview process. Thus, based upon my own reflections, it appears that interviews conducted in the research center had greater ethical integrity, as based upon Institutional Review Board standards, than some of the interviews conducted in the field. [37]

One of the benefits of conducting interviews in the field was the opportunity to conduct informal assessments of participants' living situations and their neighborhoods. The survey included several questions about satisfaction with the participants' housing and neighborhood in which they lived. Conducting the interview at the participants' homes or current place of residence provided a much richer interpretation of their responses. For example, reflecting upon an interview conducted in a rooming house, I noted:

"His building was very 'busy'. People were coming in and out and we were interrupted a couple of times. ... I'm glad we were able to do the interview at his place. It is times like this one where 'real world' experience trumps anything that can be found in a controlled environment. The interview would probably have been the same, but actually being able to experience the living situation of participants is invaluable." [38]

Similarly, after conducting an interview in a shelter, I wrote, "I think doing the interview within the walls of the shelter helped me understand a lot more of what it is like to live at this particular shelter ... and I was able to see the conditions of the rooms where people slept." Although these assessments were not necessarily included within formal analyses of the data, it provided me with an opportunity to acquire a greater understanding of the lives of the participants. For interviews conducted in the research center, I was not able to acquire such valuable experiential knowledge. [39]

Having the option of conducting interviews at the research center proved to be beneficial for participants without housing or with unstable and unsafe housing. One research participant stated that his current roommate was a "junkie" and that he sometimes found syringes lying on the ground. Another participant noted that an assault recently occurred in his building and that I would probably not have felt comfortable conducting the interview there. Other participants spoke of the lack of privacy in their apartments or rooms and the possibility of bed bugs being present. Across these examples, it appeared to me that the participants brought in their own ideas of what an appropriate research environment is and preferred the interview to be conducted at the center. In this sense, it was important to have the option to conduct interviews at the research center and highlights some of the limitations of interviews conducted in the community. [40]

Having the participants choose where they wanted the interview to take place was a lost opportunity within the initial interviews. I did not ask the participants where they would like the interview to be conducted. I believe this resulted from my lack of recognition of the power that interview locations could hold. When I began to offer participants the choice in location, they did not exclusively choose one setting more than others. [41]

Participants who chose to conduct the interview at the research center had varying rationales for this choice. For participants currently residing in the shelter system, the research center was a suitable location since it was located very close to the three main shelters in the city. Several participants noted that they were familiar with the area and could easily locate the building. Some participants were also residing quite close to the center, so travel was not an issue. [42]

A limited number of participants preferred not to travel to the center since it was located in the downtown core of the city. For participants in recovery from substance abuse, being downtown served as an area that could trigger substance use. These participants had removed themselves from the downtown core and preferred to not return, as evident in the following journal excerpt:

"He mentioned that during his first interview, that was conducted downtown, that he spent the honorarium on drugs. He also mentioned how he liked living away from downtown, as he is in recovery and doesn't want to see the old people/habits that he did when he was not in recovery." [43]

I later journaled that I did not realize the significance that interview locations could potentially have on participants until this moment. It underlined the importance of providing choice for participants when organizing the details of the interview. These kinds of reflections led me to view the participants as much more than just participants. Each individual I interviewed had a multifaceted identity and these identities needed to be respected throughout the research process, including the selection of an interview site. To neglect this would hamper the research. [44]

Some participants appeared to view the interview as a social opportunity, regardless of setting. For participants that came to the center, some stated that this was a, "chance to get out of the house" and that they enjoyed the interview. For example, in one of my journal entries I stated:

"Seems to keep to himself, by choice, but was enjoying the interview. Led me to believe that his choice of doing the interview at [the research center] was so he could get out of the house and engage in some social interaction. We set up the interview the same day, so it didn't appear that he had other plans for the day." [45]

Similarly, in one entry I wrote, "[I] went to participant's housing. Offered to do the interview at his place, but he said that it would be good for him to get out." In these cases, the research center served as a setting for which participants could leave their places of residence and engage in a new environment. In a limited number of cases, the social function of the interview interfered with the research process as some individuals provided lengthy answers and were hard to focus. [46]

For interviews conducted in the home, having a guest enter their apartment appeared to serve an important social function. For example, after one interview conducted in a participant's home, I noted, "I don't think he would have been as comfortable if the interview was conducted outside of his home. He stated he lacks social contacts ... Once the interview was done, he was really eager to chat about anything." The interview serving as a social function was another unintended outcome that I reflected upon in my journal entries. For socially isolated participants, the selection of the interview location could mean the opportunity for a social interaction that may not otherwise happen. For individuals who had experienced homelessness but were now housed, the prospect of hosting a guest may have very important in the process of being re-housed and gaining comfort in this position. [47]

My personal safety was a recurring theme within my journal entries. I noted that I was glad some interviews were conducted at the center or within community settings (e.g., shelters, community agencies) since some participants were irritable and others used mildly threatening language. During these situations, I was comfortable knowing that I could stop the interview and ask for assistance from employees of the center. Had these types of interviews been conducted in the participants' homes, I may not have felt the same sense of safety. [48]

Safety issues were also present when I conducted interviews in the homes of participants. Although the majority of situations were very positive, I felt that they did place me in potentially compromising situations. For example, during one interview I did not realize that there was someone else present in the apartment. The individual only came out once the interview was finishing. In another instance, a large dog came into the room during the interview. Luckily the dog was friendly, but I was unaware that there was an animal present. As a safeguard, I received training on interview techniques (i.e., ensuring that you are aware of exits, sitting closest to the door, etc.), sat in on several interviews by experienced interviewers, always had access to my cell phone and the study coordinator was aware of my location, so I was prepared to notify the study coordinator if I felt uncomfortable during the home interviews and required assistance. Some interviews were also conducted in pairs by me and another researcher. [49]

An unexpected theme that was identified was the hospitality of the interviewed individuals and the lack of hospitality that I was offering at the research center. When conducting interviews in the field, I noted that some form of beverage was often offered by the participant. Related to a previous point, this affirmed my role as both researcher and guest. This same kind of behavior was not initially present when I was conducting interviews at the research center. Eventually, I began to offer participants water at the outset of the interview, but I felt it was less meaningful than when a participant offered something to me. My original lack of hospitality in the research center may have set a tone for the interviews, one where I was the gatekeeper of the interview and one where I had the power. Location dictated hospitality which in turn may have impacted how participants responded to the interview questions. [50]

This introspective study highlights the importance of location when conducting research with individuals experiencing homelessness and vulnerable housing. Through the use of personal research diaries, it was found that location is an important factor to consider when conducting research interviews. These locational impacts can affect the research process, the collected data, the interviewees, and the interviewers. [51]

As evident in the data, interview sites hold immense amounts of power on the research process. The differential power distributions based upon the interview site replicates previous work with homeless and vulnerably housed populations (COHEN et al., 1993). The greater power imbalance at the research center may reflect my training as a research psychologist. Psychological research often follows a positivist framework in which researchers engage in a one-way relationship with their participants (BRODSKY, 2001). At the research center, I was in control of the research process and dictated the course of the interview. As a result, my delivery of the survey material may have been more systematic, eliciting more formal responses from the interviewees and perhaps stronger internal validity for the study. For interviews conducted in the field, this power was somewhat diminished. This could have resulted from my awareness of the multiple identities and roles that participants were exhibiting (ELWOOD & MARTIN, 2000). In the field, I was able to witness the participants as apartment dwellers, rooming house residents, shelter inhabitants, and social service users. Their roles became more expansive than that of just a research participant. [52]

Confidentiality issues only arose outside of the research center. Choosing private interview locations is a concern for homelessness research (HWANG et al., 2011). Interviews conducted in the field are subjected to external influences that can be controlled within a more formal research setting. This may affect the data collection process, as interviewers may focus their attention on potential for threats to the ethical integrity of the interview. To remedy this, it may be beneficial to scout out suggested locations prior to conducting field interviews to ensure that the location is suitable, although this may not be possible if participants set up same day interviews. [53]

Field interviews allowed for a deeper understanding of the lives of the participants than those conducted at the research center. An assessment of interview locations can lead to a deeper understanding of information discussed during the interview (ELWOOD & MARTIN, 2000). As many of the interview questions in the current study focused on the housing and neighborhood experiences of the participants, it was extremely beneficial to actually witness the environments they were assessing. The same information could not be gleaned from interviews conducted in the research center. Therefore, the external validity of the information provided during field interviews may have more merit than interviews conducted at the research center. This information can be captured during post-interview written reflections by the interviewer, much in the same way it was completed in the current study. [54]

Providing choice in determining the interview location was a missed opportunity for some of the interviews. Reflecting upon the interview choices brought forth by the participants serves an important function, since they may highlight locations important to the participant (ibid.). Interview location selection may also have more personal reasons for individuals experiencing homelessness and vulnerable housing. Self-imposed sanctions affected interview location decisions for some participants. For example, individuals in recovery from substance use found certain locations unhealthy for their recovery, such as the neighborhood our research center was located. Related to this point, given the sensitive nature of some of the questions (e.g., homeless experiences, substance use) some participants may have selected an interview site where they felt most comfortable. For other participants, the interview served as a social opportunity, whether it be at the research center or within their home. This is an important consideration given the social isolation that many marginalized group's experience and a result that replicates findings from RYAN and GOLDEN (2006). [55]

Providing choice in the interview location also serves a methodological function in that it provided an opportunity to shape the research process. This type of involvement is important as the involvement of marginal participants in the research can increase its relevancy and inclusivity and impact the hierarchical nature of research itself (TERRY, 2016). Therefore, it is imperative that choice is provided to homeless and vulnerably housed research participants. [56]

Interestingly, transportation costs were not part of my reflexive analysis. This was in part due to clients being offered bus tickets if they chose to have the interview conducted at the research center; however, for participants who chose to have the interview conducted at their homes or a location in closer proximity to their housing, this choice may have resulted from a lack of financial resources. [57]

In respecting participants' choice, researchers must also be cognizant of their own safety. As highlighted in previous research, interviewer safety is a particularly important considerations for interviews conducted in the field (COHEN et al., 1993; TORO, 2006). Therefore, the training of interviewers is especially important so that they are made aware of the potential situations that may arise during a community-based interview. Such measures as conducting interviews in pairs and having access to a cell phone should be taken when conducting interviews in the field (TORO, 2006). [58]

This study also sheds light on the importance of physical space for individuals who are homeless or who have experienced homelessness. Few studies have investigated the importance of the sociospatial dynamics of homelessness (SNOW & MULCAHY, 2001). Homelessness

"forces individuals, whose claims to community citizenship or membership are routinely contested, to continuously negotiate and survive in spatial domains of a community that were neither designed nor intended for basic resident ... and constitutes a rupture of or threat to the modern urban order" (p.154). [59]

Homeless individuals may be barred from certain public places, such as shopping centers, public institutions, or parks, as they disrupt the social order of mainstream society. Therefore, the interview location a homeless individual chooses may be in direct response to the civil sanctions that are imposed on them. [60]

This article also calls for methodological shifts in quantitative research. It is clear that many empirical articles neglect to report on the importance of the interview locations in which their studies are conducted. This is puzzling given that much psychological research is conducted in the community. Reflexivity in quantitative research, particularly that involving sensitive topics, will provide a deeper understanding of how, where, when, and by whom the data were collected (RYAN & GOLDEN, 2006). It will also benefit the researchers, as it may provide a debriefing opportunity. As well, by extending methods sections to include greater context on the interview location, richer analyses will undoubtedly follow. Reflections of location may also enhance interviewer competencies, as it certainly did for me. These reflections can convert a positivist-informed interview into a holistic analysis of a participant's life. Reflections on location can also help with a more meaningful interpretation of the data (ELWOOD & MARTIN, 2000). Techniques such as research diaries are one means to facilitate this process. [61]

This article also demonstrates that there is room for reflexivity within largely quantitative studies. Although quantitative studies often present standardized measures using forced-choice questions to participants, many participants often explain their rationale for selecting their answers particularly with regard to sensitive questions (RYAN & GOLDEN, 2006). Unfortunately, these rationales are often not recorded and this important data is lost. One method to address this lost data is for researchers to engage in reflexive processes. The reflexive accounts of the researchers in quantitative studies may provide evidence for the need to do further qualitative work with the participants of quantitative studies (RYAN & GOLDEN, 2006). It also allows for researchers of quantitative studies to note the emotions that may arise from participant responses to quantitative questions. [62]

Several limitations exist in the current study. First, only my experiences as a researcher are presented, thus there is the possibility that my personal biases impacted my research findings and interpretation of these findings. Although I did consult with another interviewer involved in the project, including other researchers in future reflexive projects would enhance the internal reliability of the findings. Second, as a male researcher interviewing primarily male participants, my experiences were limited to this same gender interaction. Female researchers interviewing male participants may dramatically differ. Related to this point, I also did not consider other defining characteristics of the participants, such as age, race, or sexual identity, which may have impacted the selection of the interview location. Third, this study is limited by the context of the city. Similar studies may highlight unique circumstances for marginalized groups depending upon the size and political landscape of the city. I did not directly ask the participants if location impacted their experience as a research participant. Last, this research serves as one means to explore the process of researching the topic of homelessness. It does not touch upon the structural causes of homelessness, such as the lack of affordable housing, insufficient income supports, and the myriad of other systemic factors which have resulted in growing homelessness rates. These structural and systemic factors should always be given consideration when discussing homelessness. [63]

In conclusion, this study has highlighted the importance of interview locations when conducting research with homeless and vulnerably housed individuals. The results demonstrate that it is important for interview locations to be given priority upon the development and implementation of community-based research projects. In doing so, the research process will be enhanced and both participants and researchers will benefit. It is also clear that each type of interview site possessed strengths and weaknesses for the collection of data. Interviews conducted in the research center benefited from methodological control, but they were not able to tap into the participant's housing and neighborhood experiences as explicitly as the field interviews. Field interviews were subject to the potential confounds, such as distractions and concern for confidentiality, but provided an opportunity for an in-depth understanding of their housing and neighborhood experiences. Recognizing the needs of the participants and being aware of safety concerns should take precedent when deciding upon an interview location, however it must be acknowledged that the chosen location may have differential impacts on interviewers, interviewees, and the collected data. [64]

Alsop, Christiane K. (2002). Home and away: Self-reflexive auto-ethnography. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(3), Art. 10, http://nbn-resolving.de/urn:nbn:de:01114-fqs0203105 [Accessed: October 26, 2016].

Brodsky, Anne E. (2001). More than epistemology: Relationships in applied research with underserved communities. Journal of Social Issues, 57(2), 323-335.

Charmaz, Kathy (2006). Constructing grounded theory: A practical guide through qualitative analysis. Thousand Oaks, CA: Sage.

Cloke, Paul; May, Jon & Johnsen, Sarah (2010). Swept up lives?: Re-envisioning the homeless city. West Sussex, UK: Wiley-Blackwell.

Cloke, Paul; Cooke, Phil; Cursons, Jenny; Milbourne, Paul & Widdowfield, Rebekah (2000). Ethics, reflexivity and research: Encounters with homeless people. Ethics, Place and Environment, 3(2), 133-154.

Cohen, Evan H.; Mowbray, Carol T.; Bybee, Deborah; Yeich, Susan; Ribisl, Kurt & Freddolino, Paul P. (1993). Tracking and follow-up methods for research on homelessness. Evaluation Review, 17(3), 331-352.

Dowling, Maura (2006). Approaches to reflexivity in qualitative research. Nurse Researcher, 13(3), 7-21.

Ellis, Carolyn; Adams, Tony E. & Bochner, Arthur P. (2011). Autoethnography: An overview. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(1), Art. 10, http://nbn-resolving.de/urn:nbn:de:0114-fqs1101108 [Accessed: October 26, 2016].

Elwood, Sarah A. & Martin, Deborah G. (2000). "Placing" interviews: Location and scales of power in qualitative research. The Professional Geographer, 52(4), 649-657.

Ensign, B. Josephine (2006). Perspectives and experiences of homeless young people. Journal of Advanced Nursing, 54(6), 647-652.

Finlay, Linda (2002). "Outing" the researcher: The provenance, process, and practice of reflexivity. Qualitative Health Research, 12, 531-545.

Hall, Wendy H. & Callery, Peter (2001). Enhancing grounded theory rigor: Incorporating reflexivity and relationality. Qualitative Health Research, 11, 257-272.

Hwang, Stephen; Aubry, Tim; Palepu, Anita; Farrell, Susan; Nisenbaum, Rosane; Hubley, Anita M.; Klodawsky, Fran; Gogosis, Evie; Hay, Elizabeth; Pidlubny, Shannon; Dowbor, Tatiana & Chambers, Catherine (2011). The health and housing in transition study: A longitudinal study of the health of homeless and vulnerably housed adults in three Canadian cities. International Journal of Public Health, 56, 609-623.

Jordan, Amy B. (2006). Make yourself at home: The social construction of research roles in family studies. Qualitative Research, 6(2), 169-185.

Kann, Laura; Brener, Nancy D.; Warren, Charles W.; Collins, Janet L. & Giovino, Gary A. (2002). An assessment of the effect of data collection setting on the prevalence of health risk behaviors among adolescents. Journal of Adolescent Health, 31(4), 327-335.

Kelly, James G. (1990). Changing contexts and the field of community psychology. American Journal of Community Psychology, 18(6), 769-792.

Mowbray, Carol T.; Cohen, Evan & Bybee, Deborah (1993). The challenge of outcome evaluation in homeless services: Engagement as an intermediate outcome measure. Evaluation and Program Planning, 16, 337-346.

Nadin, Sara & Cassell, Catherine (2006). The use of a research diary as a tool for reflexive practice: Some reflections from management research. Qualitative Research in Accounting & Management, 3(3), 208-217.

Newbrough, John Robert (1992). Community psychology in the postmodern world. Journal of Community Psychology, 20, 10-25.

Rosenthal, Rob (1991). Straighter from the source: Alternative methods of researching homelessness. Urban Anthropology and Studies of Cultural Systems and World Economic Development, 20(2), 109-126.

Ryan, Louise & Golden, Anne (2006). "Tick the box please": A reflexive approach to doing quantitative social research. Sociology, 40(6), 1191-1200.

Saldana, Johnny (2009). The coding manual for qualitative researchers. London: Sage.

Snow, David A. & Mulcahy, Michael (2001). Space, politics, and the survival strategies of the homeless. American Behavioral Scientist, 45, 149-169.

St. Louis, Kathleen & Calabrese Barton, Angela (2002). Tales from the science education crypt: A critical reflection of positionality, subjectivity, and reflexivity in research. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(3), Art. 19, http://nbn-resolving.de/urn:0114-fqs0203196 [Accessed: October 26, 2016].

Taylor, Cheryl (1993). On the conduct of homelessness research: Lessons from a qualitative study of women diagnosed with chronic mental illness. Issues in Mental Health Nursing, 14, 425-432.

Terry, Linda (2016). Refreshing perspectives: Exploring the application of peer research with populations facing severe and multiple disadvantage. London: Revolving Doors, http://vibrantcanada.ca/files/refreshing-perspectives.pdf [Accessed: October 20, 2016].

Toro, Paul A. (2006). Trials, tribulations, and occasional jubilations while conducting research with homeless children, youth, and families. Merril-Palmer Quarterly, 52(2), 343-364.

Vance, David (1995). Barriers and aids in conducting research with older homeless individuals. Psychological Reports, 76, 783-786.

Walsh, Richard T. (1987). The evolution of the research relationship in community psychology. American Journal of Community Psychology, 15(6), 773-788.

John ECKER is currently the Director of Evaluation at the Canadian Observatory on Homelessness and formerly a research associate with the Centre for Research on Educational and Community Services at the University of Ottawa, Canada. He received his PhD in experimental psychology from the University of Ottawa in 2015. His research focuses on homelessness, vulnerably housing, community mental health, community integration, and issues related to the LGBTQ community.

The Canadian Observatory on Homelessness
6th Floor Kaneff Tower, York University
4700 Keele St., Toronto, ON M3J 1P3, Canada

Ecker, John (2017). A Reflexive Inquiry on the Effect of Place on Research Interviews Conducted With Homeless and Vulnerably Housed Individuals [64 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 18(1), Art. 5,
http://nbn-resolving.de/urn:nbn:de:0114-fqs170151.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License